mmc: core: Fix recursive locking issue in CQE recovery path
authorSarthak Garg <sartgarg@codeaurora.org>
Thu, 7 May 2020 16:15:33 +0000 (21:45 +0530)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 20 May 2020 06:20:21 +0000 (08:20 +0200)
commite8eb122b9f438f30e0d867f332ab5ff72c40d294
tree201ec3bc1fbecabc182b9a827fde6128c3873145
parentfdf547a591f51964ab166ad8b05cda75a050b659
mmc: core: Fix recursive locking issue in CQE recovery path

[ Upstream commit 39a22f73744d5baee30b5f134ae2e30b668b66ed ]

Consider the following stack trace

-001|raw_spin_lock_irqsave
-002|mmc_blk_cqe_complete_rq
-003|__blk_mq_complete_request(inline)
-003|blk_mq_complete_request(rq)
-004|mmc_cqe_timed_out(inline)
-004|mmc_mq_timed_out

mmc_mq_timed_out acquires the queue_lock for the first
time. The mmc_blk_cqe_complete_rq function also tries to acquire
the same queue lock resulting in recursive locking where the task
is spinning for the same lock which it has already acquired leading
to watchdog bark.

Fix this issue with the lock only for the required critical section.

Cc: <stable@vger.kernel.org>
Fixes: 1e8e55b67030 ("mmc: block: Add CQE support")
Suggested-by: Sahitya Tummala <stummala@codeaurora.org>
Signed-off-by: Sarthak Garg <sartgarg@codeaurora.org>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/1588868135-31783-1-git-send-email-vbadigan@codeaurora.org
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
drivers/mmc/core/queue.c